Toughness Of Lead-tin-antimony Alloys

[grumpy one]

TOUGHNESS OF LEAD-TIN-ANTIMONY ALLOYS
GEOFF CHAMBERLAIN
When choosing which metal to use for a specific application, physical properties such as strength, hardness, ductility and toughness are often important. Toughness, defined as the ability to withstand shock loading, is commonly determined by the Charpy test (ASTM E23). This involves a pendulum breaking a standard specimen: the pendulum’s loss of energy from before-impact to after-impact is the toughness measurement.

Toughness is a critical characteristic used by the military to compare the impact performance of specimens of armour plate. It has the potential to be similarly useful for comparing the expected behavior of bullets cast from various lead-tin-antimony alloys when they strike animals. Because I found little actual data on the toughness of these alloys, I decided to investigate the matter experimentally. In most respects the Charpy test was suitable for my experiments but I wanted to obtain a direct indication of specimen ductility. While ductility can be inferred from the ‘instrumented’ Charpy test it is not indicated by the simple mechanical version, and the instrumented version was beyond my intended scope. I therefore departed somewhat from the Charpy concept and chose to deform my specimens by a standard amount instead of just fracturing them. This still gave me a measurement of the deformation-energy required, but also enabled me to grade specimens as ductile if they deformed without cracking, intermediate if they cracked but retained considerable strength, and brittle if they fractured before reaching the standard amount of deformation.

I developed a simple drop test device, shown in the first photograph. The large cast iron weight (partly-elevated and supported by a spring-clamp in the picture) is raised a suitable distance up the graduated slide then released to fall and strike the chisel, which rests against the side of a test specimen placed on an interrupted V block. The specimen, V block and chisel are shown in more detail in the second photograph.
7554
7544

The specimens were bullets cast in a Lyman 311466 mould: a standard commercial bullet mould readily available to other experimenters, and which easily produces large numbers of physically-identical cast specimens of suitable proportions for testing. Consistent axial location of each specimen was achieved by resting the shoulder at the end of the gas check rebate against the edge of the gap in the supporting V-block. This gap was 10 mm wide. The chisel tip was flat and 2.5 mm wide. The mass of the drop-weight could be varied between 1.5 and 6.5 kg (1.5 shown in photograph), and the drop-height could be varied from zero to 30 cm. The toughness measurement was simply equal to the potential energy of the weight when suspended at its drop-height, since all of this energy was subsequently converted to kinetic energy then absorbed in deforming the specimen. The effective anvil mass was maximized by clamping the test device in a large industrial vice bolted to a 16 mm steel bench-top.

The standard amount of deformation I applied is shown in the first photograph below. Drop-height and -mass were adjusted to give an 8.5 mm dimension across each specimen after deformation, unless fracture occurred first. The specimen shown in the calliper was classified as ductile. The second picture shows an intermediate specimen, which developed a tensile crack directly opposite the chisel. The third picture shows some brittle failures: when deformed to the 8.5 mm dimension the two parts were almost separated, and could have been broken by finger pressure.

7545
7546
7547

I investigated the effects of three variables on alloy toughness: heat-treatment, percentage of antimony, and having a low tin-to-antimony ratio versus equal amounts of tin and antimony in the alloy. Heat-treatment consisted of holding a sample for one hour at a selected temperature between 175 and 240 degrees Celsius, then water-quenching.

The first experiment involved a wheel-weight alloy of about 0.2% tin and 2% antimony, remainder lead with a minor amount of arsenic. The alloys will be described by their percentages of tin and of antimony, so I will refer to this as 0.2/2 alloy. Physical analysis was not available, so all alloys were identified indirectly by simultaneous use of three methods: calculation based on their ingredients; the alloy’s liquidus temperature; and the hardness of air-cooled samples. In the absence of physical analysis, reported compositions should be considered approximate. All specimens of each alloy were cast in a single batch from a single pot of alloy. For each hardness level a sample comprising fifteen specimens was aged at ambient temperature for two weeks after casting or heat-treating, before testing. Five specimens from each sample were hardness-tested using a Lee tester, and as many of the other ten specimens as necessary were impact-tested at various energy levels until the required deformation measurement was achieved.

The effect of heat-treatment on 0.2/2 alloy is shown by the chart below. In all of the charts that follow ductile results are shown as circles, intermediate results as triangles, and brittle results as squares. The lowest-hardness sample was always air-cooled and the highest-hardness sample was as hard as I could make that alloy by simple oven heat-treatment and water-quenching.

7549
Toughness initially increased with increasing hardness, but a peak was reached at about 19 BHN and beyond this hardness, toughness declined.

In the second experiment four fairly commonplace low-tin bullet-casting alloys having different antimony contents (0.2/2, 0.9/4, 2/6 and 2/13) were compared. Results are shown on this chart.

7548

The greatest peak toughness, 7.5 Joules, was achieved by the 4% antimony alloy. Both 2% and 6% antimony alloys had less peak toughness than this, and the 13% antimony alloy had little toughness regardless of heat-treatment. In all cases peak toughness seemed to occur at a hardness close to 19 BHN.

For the third experiment three of the four low-tin alloys were compared with their pseudo-binary equivalents. A pseudo-binary alloy of lead-tin-antimony has equal percentages of tin and antimony. In such alloys substantially all of the tin and antimony are expected to combine to form the compound SbSn, so the alloy effectively is binary, or consists of only two substances: lead and SbSn (ignoring minor amounts of arsenic that may be present). The well-known Lyman No. 2 alloy, which would be called 5/5 under the notation used here, is pseudo-binary.

The first comparison was between 0.2/2 and 2/2 alloys. Results are shown in the following chart.

7550

Up to 17 BHN there was little difference in the toughness of the two alloys. Above 17 BHN the pseudo-binary alloy was both tougher and more ductile, reaching its toughness peak of 9 Joules at about 21 BHN - slightly higher than the low-tin alloy’s 19 BHN.

The second comparison was between 0.9/4 and 4/4 alloys. Results are shown on this chart.

7551

Once again the pseudo-binary alloy’s peak toughness was greater at 9.6 Joules, and was reached at a higher hardness (23 BHN) – the peak hardness achievable for this alloy. There were insufficient data points to determine whether there was a difference in ductility between the low-tin and pseudo-binary alloys.

The third comparison was between 2/6 and 6/6 alloys. Results are shown on this chart.

7552

As in both previous instances, the pseudo-binary alloy reached greater peak toughness (8 Joules) than the low-tin alloy but this time did so at only slightly higher hardness (20 BHN). As with the 4% antimony alloys, the pseudo-binary alloy’s peak hardness was lower. The pseudo-binary alloy was the more ductile of the two.

Six main observations can be made from these results. First, appropriate heat treatment enhanced the toughness of all alloys tested except 2/13, which even when air-cooled was at the 19 BHN optimum hardness for maximizing toughness. (The toughness enhancement from heat treating the next-highest-antimony low-tin alloy, 2/6, was very small.) Second, an optimum antimony content for maximizing peak toughness seemed to exist at somewhere around 4% antimony for both low-tin and pseudo-binary alloys. Third, every pseudo-binary alloy tested demonstrated substantially (26-29%) greater peak toughness than a low-tin alloy with the same antimony content. Fourth, the pseudo-binary alloys may have been more ductile than the low-tin alloys. Fifth, increasing the antimony content may have decreased the ductility of the alloys. Sixth, each alloy’s ductility may have been decreased by heat treatment. More data would be required to confirm the last three of these points.

If the fifth and sixth observations are valid, greater hardness, whether it is attained by antimony content or heat-treatment, comes at the price of reduced ductility. However the data suggest that ductility can be increased by increasing the tin content, up to the point where tin and antimony contents are equal.



This report makes use of both theoretical and empirical information from F. D. Weaver, “Type Metal Alloys”, Journal of the Institute of Metals Vol. LVI No. 1, 1935, pages 209-240.

[runfiverun]

i am happy to see this as i have thought this for some time.

this probably should be stickied.......

[Bass Ackward]

TOUGHNESS OF LEAD-TIN-ANTIMONY ALLOYS
GEOFF CHAMBERLAIN

However the data suggest that ductility can be increased by increasing the tin content, up to the point where tin and antimony contents are equal.


Yep. This is what I found.

[Ricochet]

Very interesting!

[azduke]

It is definately interesting and thank you for doing the testing

[45 2.1]

Interesting data. It would be more interesting with several more alloys with the same tests conducted with air-cooled, water dropped and oven heat treated boolits of each. There are some startling differences between some alloys and the hardness condition when you do this.

[Ricochet]

It would also be fascinating to see what happens with quenched boolits from the same batch tested at different times as they age harden and resoften.

[grumpy one]

What I am hoping will happen is that someone will offer to analyze samples of the test alloys. I can then snail-mail sample bullets to the volunteer, and when the results arrive, post the actual analysis as part of the report. That would enable me to commence the next phase: nailing down the optimum antimony content more specifically than "in the general vicinity of 4%". If I have an accurate starting point I can vary up and down relative to that point, and hopefully find the optimum to about one quarter of a percent - I'm confident about getting it within half a percent. With an accurate fix on the right antimony content, I can then check up on whether some other regularities really exist or are illusory. Ultimately the picture might become clear enough to get really ambitious and see what a touch of copper does to the story.

Heat treatment has been the PITA part of the series of experiments, because results vary. Far too often I find that I've heat treated at two different temperatures but hardness results are almost identical, and occasionally even the reverse of what I'd expected. Since there is a two week delay between heat treating and testing, that has made the process extend over about 11 weeks so far, excluding the preliminary experiments and making the apparatus. I've also accepted BHN variances of up to 1.9 for individual samples. (Anything of 2 or over was rejected and a new sample prepared). Those high variance instances were relatively early in the series, and have been improved by better heat treatment technique, but it is still pretty rudimentary - even in May I've accepted variances of up to 0.9. I have established and posted both the air cooled and maximum heat treatment BHN numbers for all of the alloys I've tested, though a different heat treatment technique might increase the maximums a bit. The hardest individual specimen I tested (one of five specimens of 0.9/4 heat treated at 240 *C, which ended up with a variance of 1.9) only achieved 29.9 BHN.

[Ricochet]

How precise and repeatable are the BHN measurements?

[grumpy one]

Well, if you mean the actual tests I did, they were all done in the same place: by filing a flat on the side of the bullet, across the three drive bands nearest the base, and indenting the middle of the flat. I always tested five bullets of each alloy/heat treatment combination, plotted the mean hardness, and calculated the variance in hardness readings. BHN reading variances ranged from 0.3 to 1.9, with a mean of 1.0 across the 34 samples. The main issue was to ensure that the indentation wasn't too close to an edge, which messed them up royally, or observably out of round.

I'm satisfied about two things: the Lee tester is the only one which actually applies the Brinell method; and my technique was somewhat consistent and generally complied with the Brinell/Lee instructions. However that doesn't mean that Lee built my particular instrument to specification, or that you can even make accurate Brinell tests on an object as small as a 30 caliber bullet. Furthermore reading the diameter of the indentation is fraught with difficulty with the apparatus involved. I've used a "real" Brinell tester, and Lee's el cheapo device isn't in that class in numerous respects. However I mounted the Lee microscope in a kiddie plastic microscope, and drilled the light-transmission hole in its stage to allow me to fit the Lee V-block to it. That gave me a stable microscope support and a stable specimen support that I could slide around. Nevertheless I found that if I went back to a month-old specimen it was not unusual to be able to read the indentation diameter differently by one Lee graduation, which seems a lot to me. Essentially I was probably getting better at using the apparatus over time. On this project alone I made well over 200 hardness measurements, and that is more than I'd made when I started it even though I'm rather prone to measuring hardness.

[Ricochet]

What I mean is, are the results repeatable enough if you were checking a piece of the same metal to make a meaningful differentiation between two samples that are, say 5% apart in BHN (as 1 sample reading 20 and another reading 19 or 21 would be.) I'm not sure small differences like that are meaningful in shooting, anyway. I'm not arguing that they aren't, I just haven't seen anything to prove that they are significant.

Or if hardness variation is detrimental (which you would expect it would be at some point), how much does it take before you start seeing the real world shooting results change? There's fertile ground for experimentation, unless someone's already done work along that line that I haven't seen reported.

Expectations are obviously critical to whether or how much it matters. If you're a competitive benchrester, you care about things like uniformed primer pockets and flash holes, precise neck thickness and concentricity, even how cases are indexed in the chamber, but if you're basically a plinker, you couldn't care less as long as it goes BANG. Same with boolit characteristics. Some will weigh every boolit and discard any outside a narrow range. Some will spin them to check for balance. Some will shoot anything that falls out of the mould and will fit into a case. Most of us are somewhere in between and have to decide what's worth paying close attention to and what's insignificant to us for our purposes (which may vary at different times.)

[grumpy one]

Seems to me there are two points here. First, I avoided the main point of your previous question because it was too hard: how much of the variability in hardnesses that I saw was due to actual variation in hardness of individual bullets within a heat treatment batch, versus how much was due to variability in hardness measurement even if all the bullets had really been of the same hardness? Second, how accurately can we expect to measure hardness, and how much does it matter?

Regarding the first question, I noticed with the samples having the highest standard deviation of hardness, that drop test results were variable: I'd get insufficient deformation on a specimen, increase the drop height by a centimeter and try another specimen, and the deformation would be less than for the previous test instead of more. Of course the opposite happened too. This never happened with samples having a low standard deviation of hardness. So, the big standard deviations seem to have been associated with non-homogeneous samples.

Turning to the second question, I've just pointed out that the overall average standard deviation I got, 1.0 BHN numbers, is an over-estimate of the variability due to hardness testing alone. However even with a standard deviation of 1.0, the statistical tables tell us that in my tests, 68% of individual test results would be within plus or minus one BHN of the correct mean for a large sample. So, at worst my measurements were within one BHN of correct two-thirds of the time, and most likely they were a bit better than that (because we know the reported S.D. (1.0) was boosted by non-homogeneous hardnesses of bullets in the same sample). How much does one BHN matter? Probably depends on how close you are to some limit, whether that means too hard or too soft. I think of it as analagous to powder charges: we should try to design our loads so that they aren't too close to some limit. If we have found by trial and error that we need a hardness of say 18 BHN to handle our load, but above 20 BHN our bullet-to-barrel fit becomes unmanageably critical, then we have a window of only 2 BHN to work in. If we are going to use up all of that and more just on the variability of our hardness measurements, we have chosen an unworkable loading condition. Get a better barrel, or reduce the load to a point where say 16 BHN is sufficient, so we have an operating window of 4 BHN instead of 2.

I've probably tested everybody's patience with that esoteric explanation, especially since it's technically wrong: averaging the raw standard deviations of samples drawn from different populations is a no-no. I should have standardized the populations before averaging.

[beagle]

Now you have me wondering how the alloys we've been playing with made from WWs and babbit would stack up with the copper content.

Felix swears by it and I've acheived pretty high velocities in the .223 with it and no leading or degradation of accuracy past 2400 FPS like with WW alloy./beagle

[James C. Snodgrass]

I'm amazed at the variability of such small proportions can significantly effect the over all out come. I am very interested in any of your future experimentation on the matter . It seems we all have a infinite desire to achieve better results but some are much more adept at the means to read all the factors to attain these better results. Great job keep up the good work. James

[grumpy one]

Now you have me wondering how the alloys we've been playing with made from WWs and babbit would stack up with the copper content.

Felix swears by it and I've acheived pretty high velocities in the .223 with it and no leading or degradation of accuracy past 2400 FPS like with WW alloy./beagle

It is mostly felix's comments on this board that make me want to look into the copper issue, but I think proper scientific method requires that I tidy up the existing loose ends before starting a new program. I can't make any further progress at all, until or unless someone volunteers to provide actual analysis of the alloys I've tested so far (doesn't anyone know a friendly scrap merchant or metallurgy lab tech?)

Then I think I should finalize the optimum antimony content, in the process acquiring some additional test data that may be sufficient to clarify the ductility trends. After that, copper is the next issue I'd hope to work on. A key question is whether there is interaction between antimony, copper and arsenic, which is why I want to start out with the optimum amount of antimony and experiment around that point rather than blunder about in a multitude of randomized constituents.

[beagle]

I'm glad that you're thinking about the copper alloys as well. Felix got me started on this track and I'm a firm believer especially for small caliber bullets in higher velocity applications.

In fact, I think this may be an area of cast bullet alloys that even the bench resters have ignored.

I guess one reason is because the alloy would be so hard to duplicate from batch to batch.

I appreciate this study as I have always believed that an excess of antimony made bullets too brittle. Finally you're shedding some light on it.

I'll be looking forward to further posts on this./beagle

[jahela]

@grumpy one:

Do you always keep the relation Sb:Sn as 1:1?
I ask because I read in the publication of Rick Kelter at LASC.ORG that
Sn is responsible for age-softening. So he limits the tin-content at 2%.
Here it's proved that the bullets keep their strength for years, the same with WW's with a still lower tin-content. What's about age-sofftening in Sb-Sn-1:1 Alloys?

Or did I as a foreigner understand something wrong?

Dirk - from a sunny germany

[Ricochet]

Thanks for the clarifications, Grumpy! I'm really looking forward to your further reports!

As for antimony/tin ratios, looking down tables of type metal compositions it seems there are a lot of 2:1 or 3:1 ratios.

I've read that arsenic is put in shot to make it round up better while falling in the shot tower. That sounds to me like something that increases surface tension, likely not something we want a lot of in our boolits.

[felix]

Correct about arsenic. So does aluminum. Arsenic will actually make the boolit harder on its own merits, where aluminum won't in comparison. I have used either to make the boolit smaller in the shank area to make the shank small enough to fit checks. Arsenic is found in many babbits as well as enough to be had in magnum shot, whereas aluminum is found in aluminum stearate. ... felix

[grumpy one]

@grumpy one:

Do you always keep the relation Sb:Sn as 1:1?
I ask because I read in the publication of Rick Kelter at LASC.ORG that
Sn is responsible for age-softening. So he limits the tin-content at 2%.
Here it's proved that the bullets keep their strength for years, the same with WW's with a still lower tin-content. What's about age-sofftening in Sb-Sn-1:1 Alloys?

Or did I as a foreigner understand something wrong?

Dirk - from a sunny germany

Hi Dirk,

According to Weaver's paper, tin and antimony bond to each other to form the compound SbSn when both are present. If Weaver is correct, when you have both Sn and Sb, but less of the former, the Sb will preferentially bond with the Sn to form SbSn, and the surplus Sb will be in the form of Sb crystals. If the amounts are equal the Sb and Sn should ideally all be in combined form. The substance SbSn acts differently from Sb, but aside from noting it had much greater scratch resistance than Sb, Weaver didn't investigate its properties.

Hence it is incorrect to think of Sn as existing in your alloy, whenever you have less of it than you do of Sb: what you have is some SbSn and some Sb. I haven't seen the LASC article you mentioned, but it appears the author is asserting that SbSn is responsible for age-softening, not that Sn is.

At this early stage I have no information on age softening beyond the anecdotal reports in individual articles. Weaver did not investigate any kind of heat treatment of her typemetal alloys - I suspect it was completely unknown in 1935.

Working on the premise that Weaver is correct, the common bullet alloys consist of lead with various proportions of Sb and SbSn crystals present. It is already apparent from the work I've done to date that a small proportion of SbSn provides substantially greater maximum toughness than an equivalent proportion of Sb crystals. On the other hand, Sb crystals provide a greater maximum hardening capability than SbSn crystals. The whole subject of age hardening and age softening is a mystery to me at the moment; I only treat anecdotal articles as a source of rumours and clues, not of facts.

[grumpy one]

Thanks for the clarifications, Grumpy! I'm really looking forward to your further reports!

As for antimony/tin ratios, looking down tables of type metal compositions it seems there are a lot of 2:1 or 3:1 ratios.

I've read that arsenic is put in shot to make it round up better while falling in the shot tower. That sounds to me like something that increases surface tension, likely not something we want a lot of in our boolits.

Nearly all typemetals are hypereutectic on antimony: they have more than 12% antimony. Hypereutectic alloys seem to act rather differently from hypoeutectic alloys, which are what we use for making cast bullets. As Weaver pointed out, hypereutectics consistently increase in hardness with increasing tin content, whereas hypoeutectics may not in some ranges. Prior to about 1934 only the pseudo-binary eutectic (10/10 in my notation) had been discovered; the low-tin eutectic, linotype (4/12) was essentially discovered by Iwase and Aoki, then demonstrated more rigorously by Weaver. Incidentally I made up a batch of 10/10 and considered making samples for test, but it ate the liner of my Lee pot so quickly that I'd have had to ladle-pour in my WW smelting pot over a gasoline stove, and I just wasn't that interested at this stage, since I know it is way, way above the optimum alloy for toughness.

With regard to arsenic I've seen a summary article somewhere which asserted that while arsenic alone is a hardening agent, its use in most modern alloys in conjunction with antimony is because it allegedly increases the sensitivity of lead-antimony alloys to heat treatment. Looking into that is one of the things I hope to do eventually.

[felix]

Yeah, Grumpy, sounds like a good idea. My contention is any poison injected into an eutectic anything will rock the alloy off any possible eutectic saturation, unless, of course, the new alloy can become an eutectic on its own accord. Prolly not possible with lead-tin-antimony in any realistic proportions for our use. Heat treating a true eutectic seems to be a waste of time. I don't think I have ever had a completely true eutectic in my pot; seems to always have had a slush stage somewhere, albeit small. ... felix

[grumpy one]

Yes felix, it seems to me that any additional crystals in a eutectic would foul up the beautiful lamellar microstructure at best. That would be like throwing a few randomly-distributed rocks onto a perfect sand garden. Most likely it would disrupt the physical properties at least to some extent.

I haven't been able to mix a perfect eutectic either so far, because all I can tell is that the outcome is not quite eutectic: I can't tell in which direction it is off. If I knew what is in my input alloys accurately though, and if my kitchen scales are any good, I should be able to hit it purely by calculation. On the other hand it is humbling to see that Weaver in her 1935 state-of-the-art laboratory, using the highest available purity grade of every constituent, sometimes missed by a whole percentage point in the composition of her test alloys when they were checked by chemical analysis. Perhaps she'd have done better to use analysed premixed constituent alloys instead of pure metals; melting antimony then adding other metals seems like a process that could result in losses.

I'm not at all confident that arsenic issues can be properly resolved by my crude apparatus and methods. Some of the material I've seen says that trace amounts can have a major effect, so just getting the arsenic content down to say 0.1% may not be enough to make its influence negligible. However, all that is for the future, and if I continue to have no way to get actual samples analysed the whole investigation is already over anyway.

[ssn vet]

that's a very clever apparatus you've worked up there.....

but...

The Charpy V-notch toughness tester is just plain fun!

Shooting those samples all over the lab. floor is one of my fonder memories :-D

[grumpy one]

In a serious lab, rather than on my all-purpose workbench, I'd have used a pendulum impact test for toughness and either fitted an accelerometer to the pendulum, or done separate tensile tests to determine ductility by measuring the percentage elongation. However the vigorous scattering of pieces of samples that you mentioned when a true Charpy test is used, is a weakness of that test design: the kinetic energy imparted to those pieces is recorded as part of the toughness measurement. That is one of the reasons for my different approach. More importantly for our purposes on this board, it is essential that test specimens be cast, not machined, and be of a size and configuration fairly similar to cast bullets. So, if I were having all my druthers, I'd have used a modified instrumented Charpy test that used a small, cast, unmachined specimen.

I'm getting underway with a new series of experiments to refine the previous results. This requires much closer control of the composition of the test alloys, so I can make more-specific observations. That is a never-ending story of course, but in the next increment I hope to be able to pin down the optimum antimony content. At this stage the idea of getting close enough metallurgical control to be able to investigate copper and/or arsenic is still way beyond the horizon.

[leftiye]

Interesting in that it seems to bear out the traditional view of tin making lead alloys tougher. Well designed test! Seems to (in my understanding) define toughness correctly, and test it well. Many references to toughness that I've seen seemed to have it confused with hardness. Also very valuable to put a range for antimony content on peak toughness. Maybe a test with some copper?

[felix]

Leftiye, there is no good way to measure how much copper is good using home equipment because of the lack of a good flux to keep it in suspension long enough to cast "perfect" boolits. That would mean sporadic measurements for toughness and/or hardness. The only practical way to play with copper is to shoot the boolits for accuracy. Perhaps the best way is to make small lots which include copper and write exact notes about that lot. Compare boolit lots at the largest target at the longest range to ever be contemplated. Using copper suggests shooting at maximum accurate (enough) velocity with what mushrooming percentage obtained. Because copper inherently makes a boolit harder as far as we are concerned, there is good reason to reduce the antimony and arsenic to maximize mushroooming. ... felix

[ssn vet]

However the vigorous scattering of pieces of samples that you mentioned when a true Charpy test is used, is a weakness of that test design: the kinetic energy imparted to those pieces is recorded as part of the toughness measurement

Awe....

What's a little KE among friends? Not to mention the energy imparted to a lab. full of undergrad. engineering geeks.

[Fleataxi]

Grumpy one:

Wouldn't it be easier if Weaver used a smooth-sided slug for his tests?

Seems all those bands add potential fracture zones, and wouldn't provide absolute measures, only relative.

A major problem with experimenting is uncontrolled variables,and even only 1 major uncontrolled variable could invalidate the whole process.

Fleataxi

[45 2.1]

What you want is a copy of a jacketed bullet, but in lead. Very hard on the outside skin and dead soft in the middle when you water drop it. If you use a low antimony mix and find a way to increase the arsenic %, you will find something close to that.

[Lloyd Smale]

first it makes me feel good about my preference for 5050 ww/lino. Its allways done extreamly well i penetration testing. the only doubt id put to your test though is the fact that the way your stressing the bullets is in no way typical of the types of strains a bullet is going to endure penetrating an animal. Im sure a bullet hitting nose first woiuld react alltogether differntly and give completely differnt results.

[leftiye]

Boolit Master





Leftiye, there is no good way to measure how much copper is good using home equipment because of the lack of a good flux to keep it in suspension long enough to cast "perfect" boolits. That would mean sporadic measurements for toughness and/or hardness. The only practical way to play with copper is to shoot the boolits for accuracy. Perhaps the best way is to make small lots which include copper and write exact notes about that lot. Compare boolit lots at the largest target at the longest range to ever be contemplated. Using copper suggests shooting at maximum accurate (enough) velocity with what mushrooming percentage obtained. Because copper inherently makes a boolit harder as far as we are concerned, there is good reason to reduce the antimony and arsenic to maximize mushroooming. ... felix

Felix, Suspension? Is there not a percentage where a true alloy occurs? If there is, then carbon bearing fluxes should stop oxidation. If copper does not alloy, then forget it. Hell, don't they make lead bearing steel alloys?

I'm just fine with lead/tin alloys (up to 10% tin will alloy), I'm not convinced the antimony has any desireable effect on toughness anyway. If it were 10% tin, then maybe it will "carry" more copper. I know copper and tin will alloy. In fact, how about putting in copper that is already alloyed with tin - as in brass?

[grumpy one]

Grumpy one:

Wouldn't it be easier if Weaver used a smooth-sided slug for his tests?

Seems all those bands add potential fracture zones, and wouldn't provide absolute measures, only relative.

A major problem with experimenting is uncontrolled variables,and even only 1 major uncontrolled variable could invalidate the whole process.

Fleataxi

Weaver - a lady by the way - only did cooling tests and hardness tests, she didn't do any impact or strength tests. She also focused on typemetals, so she didn't go below 6% antimony (2/6 alloy, once known as electrotype in British industry, was the nearest she came to pure lead). The concept of using a bullet-sized specimen and impact testing it was mine, not hers.

I was initially attracted to the idea of using a straight smooth specimen of 7mm diameter. I chose the 311466 for practical reasons, one of which was that I wanted the face-validity that goes with testing a real bullet. Another was that any bullet caster or metallurgist can acquire the same mould and duplicate my tests if he or she wishes; the whole issue of sample preparation disappears except for the matter of mould temperature and casting technique. So far as the issue of obtaining absolute measures is concerned, I don't know of any metallurgical test of physical properties that provides anything but relative measures - especially the Charpy test, with its notched specimen. The essential feature in this case is to get the right kind of grain structure. This depends on the specimen's surface configuration, diameter, length, pouring arrangement, metal temperature and mould temperature. The arbitrary feature of my approach was the use of a 30 caliber specimen rather than say 45 or 22 caliber. Rejecting the large caliber was easy: such bullets are way too short and fat to be fractured repeatably by transverse impact. The objection to the small caliber was partly that repeatability would have suffered due to absolute scale effects (shrinkage differences with different alloys would have mattered, as would casting quality), and partly that it is at one extreme of the range of specimen sizes likely to be of interest.

[grumpy one]

What you want is a copy of a jacketed bullet, but in lead. Very hard on the outside skin and dead soft in the middle when you water drop it. If you use a low antimony mix and find a way to increase the arsenic %, you will find something close to that.

The straight smooth specimen concept has an attractive simplicity, but real cast bullets come with a range of built-in stress concentrations, so I decided to include some. Since the 311466 is a Loverin, I ended up including plenty rather than some. The alternative of two or three large lube grooves (which I tried in pilot experiments) seemed as if it might have been sensitive to longitudinal location of the impact point relative to the edges of the grooves, so my solution was just to have lots of small grooves and a fairly precise way of fixing the axial location as well.

So far as skin-versus-core hardness is concerned, my heat-treated specimens were water quenched of course, so they would have had a variation in effective quench rate versus radial depth generally similar to water dropping. The real issue here though is specimen diameter; you can't expect the same hardness distribution between surface and core in a 30 caliber bullet that you get with a 45.

In the series of tests I've just begun I'm trying to get considerably better control of alloy composition, and I hope I now have a way to get actual analysis results to include in my (eventual) report. This makes it possible to look at alloys only 0.5% apart in antimony content instead of 2% apart. (Since I now know from the first series of tests that I'm only interested in the range between about 3% antimony and just over 4%, the number of test alloys remains manageable.) Of course it does not make my variability in heat treatment or physical test parameters any less, so I don't know at this point whether the whole project will succeed this time; I may end up with a lot of wavy graphs that keep intersecting with each other. However that is the way it goes with experiments - if we knew the results we wouldn't need to perform them.

So far as arsenic is concerned, the literature seems to say that it has little incremental effect above about 0.15% content if used in conjunction with antimony. Equally importantly the analyzer I hope to be able to access does not test for arsenic. I can get a proper lab analysis for $50 per test, but at this point I'm not greatly attracted to that option, given the number of tests likely to be required, so I'm hoping to get by using X-Ray Fluorescence with one of the ubiquitous Niton XLt scanners through the cooperation of a scrap dealer.

[grumpy one]

first it makes me feel good about my preference for 5050 ww/lino. Its allways done extreamly well i penetration testing. the only doubt id put to your test though is the fact that the way your stressing the bullets is in no way typical of the types of strains a bullet is going to endure penetrating an animal. Im sure a bullet hitting nose first woiuld react alltogether differntly and give completely differnt results.

Lloyd, if my results are valid your alloy has way too much antimony to have good toughness. You can see that by comparing my 0.9/4 graphs with my 2/6 graphs; 2/6 is a bit on the brittle side, and your alloy must be 7 or 8% antimony.

There isn't any single, simple test that can duplicate all the things that might happen to a bullet under impact. What I hope to achieve is a valid, repeatable laboratory test that shows something about the metallurgical properties of these alloys. How that interacts with particular bullet designs, impact velocities, and sets of circumstances like how the bullet is oriented when it strikes bone, are a different line of inquiry. My expectation is that a bad alloy remains inferior to a good alloy regardless of how you apply it in practice - but the way you apply it may make more difference than the alloy properties do.

[Lloyd Smale]

i guess i still believe the best test that is available now for the average man for comparing how bullet alloys differ in a hunting situation is a good heavy bone followed by wet news print. If a guy is not lazy and changes his paper and bone often enough it gives a pretty good look at how a bullet of a certain alloy compares to another. The results are surpisingly repeatable. It all depends on what a guy considers toughness to be. To me its not only the bullets ablility to keep from fracturing but also has to take into count the resistance to deform. Some alloys are very ductable and resist fracturing but arent worth a hoot because they will deform apon hitting bone. Something thats fine in a whitetail but if a guy has to rely on a bullet to put down a 1000lb animal you want as little deforming as possible. I guess my point is a flat nose bullet cast out of a slightly higher antiomy content will give a better compromise when it comes to holding together and not deforming. Now another thing you have to keep in mind with my thoughts is that they are for magnum handgun velocitys and bullets not rifles. To be honest ive never seen a bullet out of straight linotype fracture at handgun velocitys. I have though seen water dropped ww bullets, especialy sharp nose to driving band designs like swcs loose there noses.

[MakeMineA10mm]

Hiya grumpyone. I got a chance to read this, so I'm posting here instead of the other thread.

First, a suggestion. Instead of mailing off your bullets to get their exact composition certified, why not order laboratory-grade lead, tin, and antimony - scale weigh each into the alloys you want to test - and cast a few sample bullets to test? Thus, you'd avoid all the issues of lab GCMS, as well as "contamination" (or maybe more accurately) variation in alloy. Just a suggestion.

Second, I'll have to re-read your posts here several more times, but on our discussion on the other thread, you (very paraphrased) said that high-tin with low-antimony alloys have a poor quality consistency of hardness. I don't doubt that assertion, but I don't see the data here to support it. (Like I said, I'm going to re-read it a couple more times - being neither an engineer or chemist, you're going pretty fast for me... [smilie=1: ) It SEEMS to me that most of Weaver's testing was done with very high concentrations of tin/antimony vs. what I was talking about with a low-BHN cowboy bullet. Could it be the conclusions with those "strong-alloy" bullets do not translate to the "weaker-alloy" bullets I was talking about? (Without any proof I assert that high-lead-content alloys display good ductility. This property may counter the fracturing/tearing action one would be worried about with the inconsistency of the alloy due to the higher tin content.)

Third, I want to compliment you on your methodology. I see the scientific method often messed up, but you've done an excellent job. I believe results would be easy to correlate and proove by others.

Fourth, you state at one point that the ideal level of antimony for hardening is around 4% (but you're not sure exactly where). Does that mean, you'd assert that the best alloy would be around 92-4-4 (depending on exactly where that antimony percentage falls)?

[grumpy one]

Hiya grumpyone. I got a chance to read this, so I'm posting here instead of the other thread.

1. First, a suggestion. Instead of mailing off your bullets to get their exact composition certified, why not order laboratory-grade lead, tin, and antimony - scale weigh each into the alloys you want to test - and cast a few sample bullets to test? Thus, you'd avoid all the issues of lab GCMS, as well as "contamination" (or maybe more accurately) variation in alloy. Just a suggestion.


2. Second, I'll have to re-read your posts here several more times, but on our discussion on the other thread, you (very paraphrased) said that high-tin with low-antimony alloys have a poor quality consistency of hardness. I don't doubt that assertion, but I don't see the data here to support it. (Like I said, I'm going to re-read it a couple more times - being neither an engineer or chemist, you're going pretty fast for me... [smilie=1: ) It SEEMS to me that most of Weaver's testing was done with very high concentrations of tin/antimony vs. what I was talking about with a low-BHN cowboy bullet. Could it be the conclusions with those "strong-alloy" bullets do not translate to the "weaker-alloy" bullets I was talking about? (Without any proof I assert that high-lead-content alloys display good ductility. This property may counter the fracturing/tearing action one would be worried about with the inconsistency of the alloy due to the higher tin content.)

3. Third, I want to compliment you on your methodology. I see the scientific method often messed up, but you've done an excellent job. I believe results would be easy to correlate and proove by others.

4. Fourth, you state at one point that the ideal level of antimony for hardening is around 4% (but you're not sure exactly where). Does that mean, you'd assert that the best alloy would be around 92-4-4 (depending on exactly where that antimony percentage falls)?

1. There are two reasons I didn't use lab grade ingredients and just mix them without verifying the outcome. First, it would be fairly expensive - I used up about 70 pounds of alloy in the experiments described so far. Second, Weaver used only the highest purity lab grade ingredients then analysed them anyway, and found deviations of 1% in some cases from what she had tried for. I think the only way to know what you've tested, is to analyze what you've tested.

2. The issue of soft spots in alloys with more tin than antimony was raised and described by Dennis Marshall in the RCBS Cast Bullet Manual. He does not explain why it is so in any terms that I understand. Metallurgically I would expect that surplus tin would go into solid solution with the lead matrix, unless the tin level were extremely high - but I'm saying that without knowing what metallurgical mechanism Marshall was expecting and reporting.

With regard to ductility of very high-lead alloys, as a starting point, pure lead is extremely ductile. Addition of fairly modest amounts of antimony or arsenic seem to reduce this ductility considerably. It is evident from my own results shown in this thread that ductility decreases considerably from 2% antimony, to 4% antimony, and then more rapidly to 6% antimony, and by 13% antimony ductility is pretty well negligible. However ductility seems to be higher for pseudo-binary alloys than it is for low-tin alloys with the same antimony level, and toughness is considerably higher for the pseudo-binary alloys. For your cowboy bullets I suggested in the other thread that you use something considerably less than 1.5/1.5 alloy. Without knowing what pressure you are using, I'd have thought pure lead would be adequate but isn't all that easy to cast, so you might consider lead with just 1% tin added.

3. Thank you.

4. I have started a series of experiments aimed at identifying the toughest-possible lead-tin-antimony alloy more precisely than just as approximately 4/4. I don't know if the experimental method I'm using will work or not; if it does, it will take a while yet. Until I've found the alloy that gives the highest peak toughness, then had it analyzed, I can't go any further than to say somewhere fairly near 4/4. It is highly frustrating not having ready access to an XLt scanner. More importantly though, I don't recommend 4/4 alloy as a general purpose bullet alloy because it is both more expensive and less ductile than say 2/2, 1.5/1.5, 1/1, and pure lead. I'm anticipating that the best solution is to only provide enough toughness to suit the energy level of your bullets, while maximizing ductility. That means use just enough alloying to heat treat to give the hardness and toughness you need for your application. That way you get the lowest cost and highest ductility together with the amount of toughness you have found you need. 4/4 alloy is my current estimate of what is the toughest possible alloy, which makes it a high-performance rifle alloy. Short of a bolt action handgun, it probably isn't needed for anything else. People who want low-energy hollow-point bullets to expand readily would probably be better-served by much lower degrees of alloying.

[felix]

You're doin' good, Grumpy! Keep it up. ... felix

[jahela]

Hello Grumpy,

I just read your answer #20 from 5-26-2008 to my question, also the rest of this thread. I wanna say thank you for your work you made for all serious casters.

Dirk

[MakeMineA10mm]

Thanks Grumpy. It all makes pretty good sense to me, now that I've re-read it and you've responded to my questions. I agree with most everything you say, especially about people making their boolits too hard.

I found what is probably the same article by Dennis Marshall in the Lyman Cast Bullet Handbook. It's about a 12 or 13 page article which goes into a lot of scientific analysis of lead and lead alloys. Talks about binary & ternary alloys, metallic solutions, etc. More light reading! :mrgreen:

[azduke]

Reading this is very interesting and makes one do some thinking.I'm new to this and am trying to learn as much as I can.Planning on casting my own when I get my project rifle put together.Was planning on using the wheel weights I have stockpiled as a basis,but according to an nra contributing writer wheel weights are to "hard" for round ball and black powder cartridge bulletts

[yodar]

Do you equate ductility with ability to obdurate?

Yodar

[grumpy one]

Do you equate ductility with ability to obdurate?

Yodar

Ductility is the extent to which a material can be deformed plastically without fracture. Thus it is extremely high for pure lead, and extremely low for most type-metals. It seems to me that this is closely related to, say, a plug's potential to be expanded permanently in diameter without cracking, by squeezing it axially. In other words, ductility should be a measure of the extent to which a bullet is capable of obturating without cracking. However it does not relate to the pressure required to make it obturate, although many shooters probably treat these two considerations as closely related. There are some more subtle factors that come into the picture too: for some materials ductility in tension and compression may be different.

[Marlin Junky]

Maybe someone has already addressed this point, but how do the larger diameter boolits respond to the toughness testing? Would you be willing to repeat the tests with .375 cal boolits? I'm sorry I didn't have time to carefully read everything, but what were your casting conditions?

MJ

[grumpy one]

Maybe someone has already addressed this point, but how do the larger diameter boolits respond to the toughness testing? Would you be willing to repeat the tests with .375 cal boolits? I'm sorry I didn't have time to carefully read everything, but what were your casting conditions?

MJ

There are two issues with larger diameter bullets - one test-related, the other metallurgy-related. The test-related issue I've discussed above: the length to diameter ratio of the bullet is important. Short fat bullets are unsuitable for transverse fracture testing. In addition there is a scale issue: the actual impact energy measurement would increase with bullet diameter, so it would be necessary to repeat the entire series of experiments, on all alloys, for each calibre. Just as the Charpy and Izod tests use a single, arbitrary specimen size to represent an alloy, I have used 30 calibre and am continuing to do so in the follow-up experiments. I do not think repeating the entire series with a different calibre would add anything to the science content of the work, except in relation to the metallurgy-related issue.

The metallurgy-related issue has been raised by several people in the discussion above and elsewhere on this board: since the quenching process removes heat only via the exterior of the bullet, the surface-to-volume ratio will influence how quickly the core of the bullet cools. Most likely large calibre bullets will show more variation in hardness from surface-to-core than will small calibre bullets. This is a subject that might be worth investigating, but not as part of a series of experiments aimed at researching toughness.

I've set out most of the details of casting conditions in the report, and supplemented it to some extent in the discussion above. I haven't mentioned that I generally followed my standard approach to casting: start with a metal temperature of 750 F until the mould warms up and the bullets become frosty, then wind back to 650 F. I use a lower temperature than 650 F for most moulds, but this series of experiments uses 311466, which is a Loverin design. My experience has been that I need perhaps 50 F higher casting temperature for this (Lyman DC) mould than I do for two-groove 30 calibre designs such as 311291. Conversely, the little 311255 is so easy to cast I use an even lower temperature - typically less than 600 F. The bullets were dropped onto a towel and air cooled, then heat-treated at a later date.